Thirty cocaine-dependent subjects were enrolled into a cue-laboratory
study to determine the specificity and sensitivity of this paradigm as
a craving measure. Subjects experienced three cue types (i.e., cocain
e, arousing, and neutral stimuli) in three cue modalities (i.e., audio
, visual, and manual). Cue types were administered in different experi
mental sessions with a period of 2 to 3 days between sessions. Our res
ults showed that subjective and physiological craving for cocaine was
relatively specific for the cocaine cue as compared with either the ar
ousing or the neutral cue. The relative sensitivities of the cue modal
ities in decreasing order was manual, audio, and visual. We suggest th
at a modified conditioned-cue paradigm could be a useful tool in the r
epeated assessment of craving during a clinical study. (C) 1998 Elsevi
er Science Ltd.